Qualitative Responses Based on a Technological Theme

a. What was the driving force that required the implementation of passive RFID?

There were a great deal of lessons learned, especially during the OEF period, that helped the Combatant Commanders recognize that the logistics systems, information flow, and system responsiveness available in current DoD information technology systems didn’t work. It did not give them the information they needed, and they didn’t know whether or when it was coming or where it was. There was a need for the standardization of identifying data and material across the DoD and commercial products coming into the DoD. This strong desire to adopt a new system and the fact that the technology had come to fruition created a good technological case for RFID implementation.

Active RFID technology was obviously one of the DoD’s best tools for visibility and moving commodities in the pipeline. Once the COCOMS saw the possibilities, J-4 was able to see the benefits; there was an overwhelming desire to take advantage of the technology. In terms of passive RFID, there was no technological need that stood out from the Services or created an overwhelming desire to adopt the technology.

b. What are the significant difficulties involved in attempting to deploy a technology like passive RFID in an organization like the DoD?

Extrapolating what the business case is going to look like for passive RFID has been one of the biggest difficulties. In the Department of Defense, the Services require a business case before initiating a new program. For active RFID, there was no need for a Business Case because the technology immediately satisfied the COCOM requirement agreed upon throughout the DoD. Yet, as is typical with a new technology, there is not a lot of implementation experience regarding passive RFID upon which to reflect. Therefore, estimated savings in inventory and labor must be based on anecdotal information from a couple of small implementations. Thus, since there is not a lot of experience to draw upon to build a business case for passive RFID, the catch 22 that presents itself is which comes first: the experience or the business case?

There have also been some technical difficulties in the development of the standards for RFID tagging. Immediately after the mandate came out, even though Wal-Mart was heavily integrating RFID, connecting with the tag manufacturers was extremely difficult.

Now there are numerous upstart companies manufacturing tags, and they are seemingly all RFID experts.

It is critical to note that the entire industry may possibly have taken the wrong approach. The primary item that instigated industry investment into passive RFID was the promise of the nickel tag. It was that push for the low-cost tag that catalyzed the creation of tags of poor reliability and very poor performance. If the industry customer had been able to set tag parameters and offer contracts based on accomplished capabilities, a $1.50 tag with a read distance of 15 feet to 20 feet at all angles with a high degree of reliability would have fulfilled capabilities closer to the customer’s need. If such parameters were met, there would be a good case for the industry to create a contract for 6 billion tags and drive the cost down to 22 cents a piece. Jumpstarting an industry by generating a cheap product results in a poor product; this impacts the industry by creating a lack of interest among the users of the technology. The capability needs to perform, but it is going to cost money initially with savings down the road.

The opposing argument, of course, is that if there was no vision of a nickel tag, there would have been no incentive to adopt the technology and instigate production moving towards a low-cost tag. However, the conditioning of Defense personnel to cost sensitivity (as opposed to performance parameters) creates an expense regardless of perception.

Other problematic circumstances have surrounded the integration of the passive tags. For instance, currently, if a box with an RFID tag is placed in front of a reader, the reader will continue to read it. This would seem to be a positive result for on-shelf monitoring and warehouse management, but unless there is some sort of filter that says, “It has been read,” this multi-read glitch will overburden and bury even the most capable system. There must be reader logic that says, “Don’t keep passing this through the system” because it has seen that tag already, and it is a unique tag. The logistics system should not have to propagate that setting. The reader should be able to shut that down.

Another problem was illustrated by the Air Force. The Service did a demonstration moving RFID-enabled shipping labels through Dover. They found that the machine read ten times the number of labels the pallet required because it was tagged by someone else somewhere down the line. This creates a challenge because there were no discriminators to distinguish the required transaction from an inadvertent transaction. The Services are currently trying to develop some middleware that will handle duplicate reads and filter the reader through tag uniqueness. The DoD is working with Paul Brinkley to create architecture across the business systems in the Department to develop that capability. However, the problem is that middleware doesn’t get into the logic of the system. It will help in receipt, in receiving an advanced shipping notice, and maybe in overcoming some of the communication links, but it won’t get beyond the receipt level.

On the active side of the RFID system, there are technical issues related to the width of indiscriminate tags available that should be avoided with passive tagging. Active was initially implemented with one million serial numbers. That number was easily exceeded last year (exceeding coding capability) and required a technical workaround.

Knowing that there is a strong possibility that unforeseen barriers will develop due to the many changes required in the technological and business processes, there is a strong requirement to start collecting data on performance parameters and the “as-is” process. There is collaboration to quantify items and collect data for the development of metrics. The antennas, the way they are positioned, and interference are all places for metrics to be collected—including metrics for the business processes. The amount of data being sent either via the advance shipping notice or the data on the tag itself requires processing, discrimination, and the infrastructure to process those requirements. In order to fulfill the data-gathering needs, there must be some sort of infrastructure in place.

Once the infrastructure is positioned on specific warehouse doors, conveyors, and packaging stations, the discovery phase begins. A light switch might affect a reader; an electric box, cell phone, forklift speed, or something else could cause problems. The learning curve is exceptionally steep and costly at the front-end of an emerging technology. And in this case, that

learning curve coupled with a short time requirement means the only possible way to achieve results is to increase manpower and to achieve an impossible goal.

Once the infrastructure is in place to collect localized data, there is one additional piece that must be added before full metrics can be collected. The piece that is missing from the puzzle at this point is the overarching data system in the back that creates the visibility. Hanging the RFID out there and running it through an RFID server grants visibility from where and when it was shipped from, to in-transit information, to where and when it arrived. Right now, the legacy systems cannot determine or express that information. The problem might inadvertently be that some of the Chief Information Officer (CIO) Logistics (N-6) types are a barrier to progress. The CIOs are unable to make changes to their systems because of a lack of compatibility, or the systems are so overused that any additional system requirement would slow it down until it became useless. This is a primary result of the system’s issues.

All of the Services are in the process of rolling out an ERP system; this creates a reasonable resistance to RFID implementation because of the payback value achieved in an infrastructure update that will be replaced in four years. There is a huge infrastructure requirement for passive RFID. The data systems also need modifications. Additionally, the legacy systems were not designed to communicate with each other the way they are now expected. In order for operators to take advantage of RFID and to associate that RFID tag with the 700 legacy logistics systems available, the code of most of the systems (if not all) must be changed. That requires building up business logic to handle RFID associations or developing an interface middleware to handle the requirement. Several different legacy systems are not currently being funded because they have been “browned out” or are going to be replaced by ERPs. This problem infiltrates every Service and agency within the DoD; these are all costly endeavors to precede the installation of an ERP system.

As opposed to active RFID, there is an intuitive complexity that comes with passive RFID. There are accessible ways to use the current status of active RFID within the logic of our automation systems because it can function primarily by itself, and needed data can be pulled as necessary. But if the requirement moves towards greater granularity, there is a need to

build that capability into the logic of automation systems. This addition requires a significant investment inmore infrastructure.

These scenarios start to raise clear business questions about whether it is prudent for the Department of Defense to try to change existing legacy systems or embed the technology into the future system. The emergence of the technology and the obstacles to overcome (including cost, security, and equipment issuance) require much more infrastructure than is presently available.

Additionally, the electronic product code format brings something uniquely different and new to the table. When the electronic product code is introduced (as opposed to the standard UPC format that the systems were designed around), there is a requirement to change the entire system and logic structure, which increases the cost and timeframe of implementing passive RFID exponentially.

Besides the difficulties caused by new EPC formats, the Navy and other services have not yet conquered the two-dimensional bar codes that the DLA uses on their military shipping labels. This is a high-capacity data medium that the services were attempting to overcome when the RFID implementation took precedence. But, instead of having the experience of overcoming and applying those lessons to an RFID implementation, the only knowledge gained was that the high-capacity data medium would be a system challenge. There is a strong need for a data-management construct for the DoD, a joining logic system for the disparate legacy systems, and field-capable handheld scanners instead of field portals.

Military logistics systems are built around the fact that there are not enough secured communication networks to use as a logistics resource. With RFID, there is now a requirement to have secured communication to make passive tags work. To receipt with a passive tag, the customer must be able to receive an advance shipping notice. The soldier in the field currently gets a status code confirming satisfied requisitions and shipment departure. The soldier does not need the 80-card column with all the MILSTRIP information behind it, but the 856-S format of the advanced shipment notice sends all that and more. This elicits questions of band-width and field processing. Some of the data that the DLA cares about on a 2-dimensional bar code does not mean anything at the SYSCOM level. Yet, many system owners don’t want to

spend the extra storage money to re-create a Service-specific label in their system. With the passive tag technology, there will be a forced overflow of data because of the advanced shipping notice. In order to process that information, a system must store it in a data base or engage with some sort of middleware that can correlate the ASN to the tag when the material comes in the door.

Another aspect of the implementation difficulty is that different frequencies of operation between industry stakeholders have become a touch-point problem. For instance, the automotive and tire industry are using 13.45 megahertz communications in the high-frequency range. The OSD policy is an ultrahigh frequency at 860-915 megahertz. So, if a warehouse receives tires from the automotive industry, it will have to read both the automotive tags and the DoD-mandated tags.

There is also an overseas frequency challenge. Because the source frequencies and the power levels are different from those in the states, there are separate test requirements overseas. But the bigger challenge is getting frequency approval from the host nations. This was an issue for the active implementation, so some lessons learned are available from which to glean information. The countries who adapted the active technology are being a little friendlier than at first, but there are still issues with China and Japan who have not bought into the standard.

The 856-S ASN format is not just going to transmit tag ID information; it is going to transmit the entire data file that goes with every single tag, with a description of every single item that is inside a tagged box, in a tagged case, on a tagged pallet, etc. In the Navy, the way this format is supposed to work is the supplier sends an 856 pass—an advanced shipping notice (ASN)—that has all the information about the shipment. It has the NIIN, FSC, the cage code, and all the statistics that would have been included on the UID or the serial number, etc. The significant part is that the advanced shipping notice sends all the information from the system that is shipping the material to the system that is receiving the material. Then, the tag helps correlate the advanced shipping notice received to the one that matches the box that was just read.

In order to work the business process amongst all the Services and agencies, there can be no innovation in tag selection. If things are going to filter to the services from a DLA

warehouse, the Services have to be able to read a DLA tag; they have to be able to deal with a DLA-generated advanced shipping notice. Initially, there was a great deal of hype that all information would be put on the tag. Now, due to the inability of passive technology to read a large amount of data at a reasonable pass speed, all the information must be placed on the active tag with just enough information on the passive tag to get linked up to a previous transaction (which should have been received at that AIS via an ASN).

Since the passive tag does not have any significant information on it with regard to the shipment itself, a cynic would say that portion of the RFID is unnecessary; all that is needed is a bar code big enough to have enough digits so there is no duplication. There are several different types of 856s and several different ways to implement it to fit the business process. If each business process is different for each Service, and they are all implemented in a way that fits the Service’s legacy system, then substantial gains regarding the implementation of RFID have not been found. However, in order to achieve the potential gains in the future that most of the commercial sector is trying to capitalize on now, it is important for the DoD to find a place for the development of passive RFID as a staple technology.

There are also problematic areas as a result of the environment specific to military operations. As mentioned earlier, the shipboard environment poses a challenge because the structure reflects radio frequencies very well and is consistently subject to external high-power frequencies from radar and communications systems. The munitions environment is a challenge due to HERO (Hazards of Electromagnetic Radiation on Ordinance) concerns. There is also the combat environment which places sensitive RFID devices in the middle of the desert as opposed to in the commercial environment. Those concerns are being bounced back to the industry for analysis.

c. In your opinion, what is the current status of passive RFID deployment and investment?

The services and the DLA are labeling cases and pallets with passive tags, but the case’s life ends as soon as the case is ripped open. The Active RFID server is a separate system.

There’s no passive server in the pipeline right now. When a case gets on a truck, since there is a bar code, the bar code gets scanned. The barcode is linked to the satellite tracking unit on the

truck, and customers can drill down to the item at the NSN level on the truck coming and know its location on the battlefield. That is an active application using current bar-coding technology.

Of all the performance parameters that are available at this phase of the implementation, there is a strong requirement to achieve the capability of item-level tagging to fully benefit from passive RFID. The Navy is trying to establish a site at the IMF in Bangor, Washington. In Bangor, they currently receive a pallet, break it down, and throw a carton on a conveyer line. This carton proceeds through a portal that reads the passive tag. Hopefully, the carton on the conveyer line rolls right through. When everything works right, the passive RFID system will most likely be used. However, enthusiasm for passive RFID is tempered currently because the technology cannot do everything it was expected and advertised to do. It is very difficult for any organization, especially a public-sector Department, to take a leap of faith into the future with a technology that doesn’t work 100% of the time and has technological challenges.

The technology as yet is very immature, and the DoD is as much on the leading edge of development as anyone. There is a lot to read about what Wal-Mart and some of the commercial firms are doing, but the DoD was given the chance to see one of Wal-Mart’s warehouses. That company is not any further along than the Department is.

Soon, though, tag manufacturers profess to have the passive tags perfected. There are many makers of the new Gen 1 passive tags, and there will be many more in Gen 2. In the early stages of Gen 1, there were two main makers and several resellers of the chip. Those resellers applied their own special antennas, backings, etc. On the Gen 2 tag, there are specific chip makers and antenna makers. Therefore, the number of competing manufacturers is going to rise exponentially as we roll into Gen 2.

The Services are not conducting R&D on RFID, and they should not anticipate

The Services are not conducting R&D on RFID, and they should not anticipate